Do you mean drought?!

Dithering!

Yeah if you have a look they’re probably hollow inside. If you prune the outer branches back you’ll reveal holes into the brown, dry interior.

I’ve tried hardening off with a day release in the garden but tbh I’m just going through the motions as I can’t really see any difference! I don’t have a cold frame but I tried a poly tunnel and things just end up bolting (presumably they’re too warm for the amount of sun in spring).

I’ve been wondering about grow lights in any case, that’s worth a try. Good point about the variety, I’d been using the same seeds and maybe they just weren’t right.

Thanks for the suggestions.

I’ve sown them indoors from seed 3 years running but they invariably died after planting out. I tried various locations.

It’s possible I was a bit eager. There’s still a risk of frost until May here but I prefer to get going in April. Since they overwinter I assumed it was safe but that might only be established plants.

Our soil is really sandy without much organic matter and poor moisture retension. The beans need watering daily in summer. We’re with facing though so get plenty of sun. I may have watered inconsistently but I reckon I’ve generally got a good feel for that.

This year I got a seedling and it’s done better (is still alive). I put a lot of compost into the bed last winter and was really on it with the aphids in spring (of which there were many). Maybe we’ll see it flower next year…

They just wither and die without ever really getting established. The outer leaves seem to go brown faster than new inner ones can replace them.

I haven’t been fertilising them as they’ve always been so small. Maybe I should try that

Any tips for growing them? I’ve been trying for years without success.

This is the first time I’ve made it this far into the summer after being pretty thorough with the aphids/ ants! have a few leaves but no signs of flowering yet…

Vibe coding is accepting changes from the model without reviewing them fully (or trusting models to do the verification).

She/he is presumably discussing ideas with Claude then reviewing any suggestions before integrating them.

This may still not be to your taste but it’s not the same as vibe coding. Whether or not they are actually reviewing everything is also a separate question!

Yeah the intelligence is still in the model. The promise of symbolic AI is about logic programming/ formal semantics not recursive loops.

To a large extent the idea has failed because it proved too hard to get non-experts to represent systems formally.

I still think there’s potential value in a hybrid approach - e.g. get language models to do the representation then let them use formal reasoners/ verification instead of hallucinating.

That sounds about right to me! Trying to go full steam ahead for 8 hours a day isn’t sustainable or even desirable.

I found Cal Newport’s book Slow Productivity really helped me get some perspective on work. It has three main lessons: Do Fewer Things, Work at a Natural Pace and Obsess over Quality.

That’s not necessarily true either. You can make inference deterministic by setting the RNG seed and having the model select the highest likelihood token/ sequence instead of choosing at random.

That’s demonstrably incorrect. Different instructions give different outcomes.

Perhaps not fine-tuning weights but they can certainly use feedback from PR comments to optimise guardrails etc. And it’s definitely possible to have an agent experiment with that. And yes the Devs are the guinea pigs in that context!

You don’t need to train a foundation model from scratch. You can fine tune an existing coding model or only a LORA layer on a budget.

You could also just use the feedback to optimise system prompts, skills/ guidance or verification tools.

I presume they’re gathering this as training data and hope to later get rid of OP.

That’s why cursor went for $60bn. Anyone can make a harness. The real value is that it gathers data on which changes are accepted/ integrated.

Can’t believe the strength of opinion on this post! I assume it’s people frustrated that you dare post a pun in a (non-American) dialect… or perhaps you’ve touched a raw nerve in Oldham!

Ah I see. I’ve only played Mindustry (2D Java/ Android equivalent) and if I left that running overnight I’d wake up to complete devastation!

Hmm… I was going off Wikipedia which says 2020 - I guess the earlier release was an alpha.

In any case having your factories running in the background (overnight?) would quickly bump up those hours. Not sure it’d make for a very interesting stream though!

I never really got into quake in the same way but I did enjoy some quake 2 mods. I’d gone to uni by then so didn’t have as much free time to while away!

Tbh I’m not sure I even played through all the WADs. I got really hooked on modding which was endless fun.